238 
Fishery Bulletin 95(2), 1997 
log length (mm) 
Figure 2 
(A) Linear regression of log fecundity vs. log length in gravid 
English sole from Port Susan, Sinclair Inlet, Eagle Harbor, and 
the Duwamish Waterway. Multiple regression analysis indicated 
that fecundity was significantly higher in fish from the Duwamish 
Waterway than in fish of comparable size from the other sites 
(<=3.601, P=0.0005). (B) Linear regression of log fecundity vs. log 
age (in years) in gravid English sole from Port Susan, Sinclair Inlet, 
Eagle Harbor, and the Duwamish Waterway. Multiple regression 
analysis indicated that fecundity was significantly higher for a given 
age in fish from the Duwamish Waterway (<=6.23, P=0.0001) and 
Sinclair Inlet (t=4.33, P=0.0001) than in animals from other sites. 
(C) Linear regression of egg weight vs. GSI in gravid English sole 
from Port Susan, Sinclair Inlet, Eagle Harbor, and the Duwamish 
Waterway. Multiple regression analysis indicated that egg weight 
was significantly lower for given GSI in fish from the Duwamish 
Waterway than in animals from other sites (<=-3.218, P=0.0018). 
mals collected from the other sites; fish age (as 
estimated from length) was significantly lower 
in fish from the Duwamish Waterway and Port 
Susan than in fish from Sinclair Inlet and Eagle 
Harbor. No significant intersite differences were 
found in either condition factor or length-weight 
relationship. In contrast, other indicators of 
nutritional status showed significant intersite 
differences. Duwamish Waterway fish had sig- 
nificantly higher HSI than fish from other sites, 
as well as significantly higher triglyceride lev- 
els in plasma. Plasma glucose levels, on the 
other hand, were significantly lower in Eagle 
Harbor fish than in those from the other sam- 
pling sites. Moreover, condition factor and the 
other proposed indicators of nutritional status 
were not consistently correlated. A significant 
positive correlation was found between condi- 
tion factor and plasma triglyceride concentra- 
tions (r=0.312, P=0.014, n-6 2), but no signifi- 
cant relationship was found between condition 
factor and either HSI (r=0.093, P= 0.356, <z=100) 
or plasma glucose concentrations (r=0.036, 
P= 0.773, n=65). 
Reproductive indicators Mean values (± SE) 
of GSI, plasma 17-/1 estradiol, and plasma ALP 
(vitellogenin) for sole collected from the four 
sampling sites are shown in Table 5, along with 
results of 2-way AN OVA examining the effects 
of site and sampling time on GSI and plasma 
estradiol concentrations, and 1-way ANOVA 
examining the effects of sampling site on plasma 
ALP (measured in December only). Both GSI 
and plasma estradiol concentrations increased 
significantly between December and January 
at all sites; no significant month-site interac- 
tions were observed. Site of capture also influ- 
enced both GSI and plasma estradiol concen- 
trations. Fish from Eagle Harbor exhibited sig- 
nificantly lower GSI (£=-2.566, P=0.0115) 
and Duwamish Waterway fish exhibited signifi- 
cantly lower plasma estradiol concentrations 
(£=2.464, P=0.0156) than fish from the Port 
Susan reference site. Plasma ALP concentra- 
tions were significantly lower in sole from 
Sinclair Inlet (£=-3.004, P=0.0058) and higher 
in sole from Eagle Harbor (£=2.860, P=0.0039) 
than in fish from the Port Susan reference site. 
Ovarian atresia Ovarian atresia was also as- 
sessed in fish sampled for fecundity analysis. 
The prevalence of atresia of yolked oocytes was 
significantly higher in sole from Eagle Harbor 
than in sole from the other sampling sites ( G - 
